Diagnosing machine motion error

[maci]

Hi,

I have built from scratch a mini CNC router/robot using the MASSO G2 controller and 3 Teknic SDSK 2310S-RQN motors for the X, Y, and Z axis. After calibrating the machine, auto tuning all three motors, and running a 100 cycle positional accuracy program, I thought I had dialed in every aspect of my machine for full time use.

I am now experiencing a somewhat random motion error with my setup that I cannot seem to nail down. I am trying to use my machine as a precise dispensing robot for a very thick silicone compound and I have been running a simple program to dial in my speeds and feeds. Sporadically, at the end of the program, the machine will move the Z axis in the wrong direction while displaying that MASSO thinks it is in the right location.

I should add that when I run this code, I will manually jog the machine 0.300in in the positive Y direction to put down a new bead of silicone. I will return to the F2 screen, hit rewind to make sure the controller knows to start the program from the current location, and then hit cycle start. The error I am describing can happen whether or not I have manually jogged the machine in +Y and will sometimes repeat every cycle where the Z axis travels higher and higher, or simply remains at its new location.

The code is as follows:

%
G54
G01 Z-7.85
M03 S18
G4 P3005
M03 S10
G01 X10
M04 S55
X11
Z-7.5 M05
X0
Z-7.85
M30
%



I know this isn't best practices for a program. I only need a single line that I can repeat over and over again. The error only happens after line 10 (Z-7.5 M05) where the Z axis retracts. When the Z axis is supposed to return to the height of Z-7.85, it retracts higher instead of dropping back down towards the work piece. The display reads a value of Z-7.85 even though it has clearly moved higher to -7.15. What is more troubling is when this error occurs, it can also affect homing. I have told MASSO to home and re-find itself after this error only for it to "pull off" in the wrong direction and ram into the homing sensors. The ClearPath motor over-torque limit has saved me more times than I like to admit. Rebooting the Controller fixes this issue inconsistently.

I have monitored the ClearPath motors to see what signal they are receiving/reading when the error does happen. From what the motors can tell me, they are receiving a command from MASSO to move in the +Z direction even though MASSO thinks it is moving in the -Z direction.

Monitoring the signal voltage coming from MASSO, I am seeing fluctuations in voltage value on the +D and -D terminals when the direction error occurs, instead of a steady state signal used to indicate direction of motion.

Previously I thought I had cleared the issue by reducing the PPR in Masso and the ClearPath motors to 3200 from 6400 and re-tuning the Z axis motor. I was able to run my test program over 20 times over multiple days before seeing this occur again today.

I have hard limits turned off. Software version is "Mill 3-Axis v3.41pre"

Attached is a copy of my current settings file. Any help is appreciated.

 

[masso-support]

Hi Maci,

Could i just ask you to check the connection between Masso and the direction pin of your stepper drive please. Make sure it is tight.

If you have a bad connection the direction might not reverse intermittently.

At some stage you might want to update your software as well. The current version is 3.43 which includes a new M30 feature for cycling programs like yours.

Cheers

Peter

 

[maci]

Hi Peter,

All connections and leads checked. No loose or partial contacts found that could be causing the voltage to fluctuate at the controller.

Edit: I've also reached out to support for a new software version. I'll flash that today and see if that clears out my problem.

 

[evermech]

@maci

i've noticed in your program that you have no feed speeds programmed for the G01 linear moves. I'm wondering if you are actually rapiding those moves and losing steps in the process. Maybe you could put in an f xxx. feed speed at the first G01 move. It will use that same feed speed for all of the other linear moves unless you change to a new feed speed. Or alternately you could override your feed or rapid or both to a lower speed and see if it continues to lose its position. Just a thought.



Guy

 

[maci]

Update after flashing new software (ver 3.43): no success.

I ran the above program as a quick test without any edits to the code as a base line. After successfully homing the machine the first run of the program went off without a hitch (even seemed smoother than before). The second run of the program resulted in the machine moving in the wrong direction at Z-7.85 retract. I re-homed the machine, ran the program again and had the same result.

After this run when I homed the machine the Z axis attempted to run past the optical limit switch in the move off portion. I hit the E-stop, removed power from the z axis and manually let the axis fall. I then re-powered the z-axis and attempted a second homing sequence only for the axis to again attempt to run past the optical sensor in the move off motion (after triggering the sensor).

I then turned off the machine, unplugged both MASSO and Clearpath power supplies to clear the electronics, re-plugged both supplies back in and turned on the power. I re-homed the machine with no problem after re-setting the E-stop. I loaded the above code with a modified F command per @evermech 's suggestion only for the incorrect re-tract to happen again on the first run.



EDIT: Thanks by the way for taking the time to help me work through this.

 

[maci]

@masso-support Hi Peter, I have a quick question regarding the USB ports on the G2 controllers. Specifically if there are known issues with USB extensions or if the USB stick and Keyboar+mouse dongle are accidentally swapped.

I have all my electrical components in a steel enclosure that ideally should stay closed for the majority of the machines life. Because of this I added a dual port USB 3.0 extension to the MASSO. I grabbed one off Amazon that has dual leads so one port on the extension goes to one port on the MASSO. I have had the dongle and USB stick backwards before but I have not seen this be key to my motion error happening or not happening.

Thanks for any info you can pass along!

 

[masso-support]

Hi MACI

To be honest I don't know if wapping them causes an issue but I imagine not.

The main thing is to have the flash drive by itself and the keyboard mouse and anything else in the 2nd port on a USB hub. Don't put the Flashdrive in a usb hub.

I use a usb extension cable like you on my setup for the flash drive and it gives no problem.

With your Gcode for tidyness sake and because it should be seperate, move the M5 to a line of it's own, not that it is the cause of the problem.

What voltage do you read between the D+ & D-? Move the axis in one direction and read the voltage and then move it in the opposite direction and read the voltage again. I think one will probably read lower than the other if my understanding of the drives interface is right.

Cheers

Peter

 

[maci]

Hi Peter,

Here are my voltage readings across D+ and D-

-Z travel (down) reads about 3.55v

+Z travel reads about .125v

These are consistent with what I was reading before. This time I also checked my X and Y axis across D+ and D- and found that X and Y switch from -3.5v to +3.5 depending on the direction. I'm thinking I need to figure out what fried on the z axis controls and what fried it.

 

[maci]

Update from testing today.

I was extra suspicious about diagnosing what my electrical problem actually was and went poking and prodding with my multi-meter. What I found out was my G2 unit is not in fact damaged. I had a faulty wiring harness further down the line that was somehow preventing the voltages from completely switching state. I ended up completely replacing the harness and connection and now have what I believe to be correct voltages across all three axis.

Peter, Thanks for your patience and working with me! Hopefully that's the end of my electrical/wiring issues.